home *** CD-ROM | disk | FTP | other *** search
Text File | 1989-06-26 | 2.2 KB | 51 lines | [TEXT/GEOL] |
- Item 0108532 22-June-89 10:12
-
- From: D2086 Efficient Field Svc, C Faith, PRT
-
- To: MACAPP.TECH$ MACAPP Tech
-
- Sub: Docs and Obj pascal
-
- Larry, Keith and Jeff,
-
- I had thought of using another method which is why I asked the questions about
- the efficiency of the intra-Object calls. I think Larry's suggestion about
- having the subclass OVERRIDE the method that is called rather than calling the
- method of the abstract class is exactly what I want.
-
- I does look like it would be far more efficient than checking to see where the
- call originated manually.
-
- As far as Tdocument, Here is a scenario:
-
- I have a TDatabase document that I want to be able to handle creation of any of
- its windows. Only one TDatabase document is allowed to be open at any time. I
- want to be able to have other documents (letters for example) open as well.
- Currently If I have all the menu commands handled by the TDatabase I lose
- control when another document window is at front. Say I have one list open I
- want to get another up. I close the list window. Now since there was a letter
- open already that letter document gains control. I cannot open the new list.
- (Not without jumping through hoops).
-
- I realize that just setting the target to the document does not solve the
- situation. One must be able to have a document always gain access to commands.
- I agree this can be dangerous. But under the scenario I described I don't
- think any user interface standards would be violated by having the menus
- enabled when the letter document is at front. Having to click on a window of
- the TDatabase object to gain access to those commands is entirely too modal and
- messy. And from context none of the menu commands apply to the letter
- document, so the user would not be confused by having them active. Having the
- application do everything is also too messy.
-
- The ability to have a windowless document is not a must. I think there is
- considerable merit to Larry's suggestion that it would be nice for the user to
- have some clue that a document is indeed open. I do think this should be up to
- the programmer however, especially since it would be relatively easy to
- implement.
-
- regards,
-
- Curtis Faith
-
-
-